BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:IC Colloquium : Understanding and simplifying deep learning
DTSTART:20170330T101500
DTEND:20170330T113000
DTSTAMP:20260407T025749Z
UID:d421e325cad1dd7bb59823fd7b5550928619a3c280681557582b897a
CATEGORIES:Conferences - Seminars
DESCRIPTION:By : Armand Joulin - Facebook Artificial Intelligence Research
 \nIC Faculty candidate\n\nAbstract :\nRecently\, deep learning has been ve
 ry successful in many applications. However these approaches are often com
 putationally expensive and do not scale to the tremendous amount of data a
 vailable nowadays. It also often works as a black box making its results h
 ard to interpret for the practitioner. In this talk\, I will discuss some 
 of the recent efforts in understanding the performance and limitations of 
 deep learning based methods as well as how this insight can help to design
  simpler and faster solution to challenging problems.\n\nBio :\nArmand is 
 a research scientist in the Facebook AI Research lab. Prior to this positi
 on\, he was a postdoctoral fellow at Stanford University working with Fei-
 Fei Li and Daphne Koller. He did his PhD in INRIA and Ecole Normale Superi
 eure\, under the supervision of Francis Bach and Jean Ponce. He did his un
 dergrad at Ecole Polytechnique. His subjects of interest are machine learn
 ing\, computer vision and natural language processing.\n\nMore information
LOCATION:BC 420 https://plan.epfl.ch/?room==BC%20420
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR